Absolute 3D hand pose and shape reconstruction from a single head-mounted camera is essential for smart-glasses-based AR, telepresence, and hand-centric manipulation. However, monocular RGB methods suffer from depth–scale ambiguity and poor generalization across diverse head-mounted camera models, often requiring costly device-specific training data. We introduce EgoForce, a unified monocular 3D hand reconstruction framework that recovers robust camera-space hand pose and position across fisheye, perspective, and distorted wide-FOV cameras. EgoForce combines a differentiable forearm representation, a unified arm–hand transformer, and a ray-space closed-form solver to stabilize pose estimation and resolve absolute 3D geometry across camera models.</p>\n","updatedAt":"2026-05-13T20:01:34.000Z","author":{"_id":"641555d9385a75d7790fb343","avatarUrl":"/avatars/255aec71e94f380de5af809c766404b7.svg","fullname":"Christen Milller","name":"chris10","type":"user","isPro":true,"isHf":false,"isHfAdmin":false,"isMod":false,"isUserFollowing":false}},"numEdits":0,"identifiedLanguage":{"language":"en","probability":0.8660178780555725},"editors":["chris10"],"editorAvatarUrls":["/avatars/255aec71e94f380de5af809c766404b7.svg"],"reactions":[],"isReport":false}}],"primaryEmailConfirmed":false,"paper":{"id":"2605.12498","authors":[{"_id":"6a04bcd5b1a8cbabc9f0856d","name":"Christen Millerdurai","hidden":false},{"_id":"6a04bcd5b1a8cbabc9f0856e","name":"Shaoxiang Wang","hidden":false},{"_id":"6a04bcd5b1a8cbabc9f0856f","name":"Yaxu Xie","hidden":false},{"_id":"6a04bcd5b1a8cbabc9f08570","name":"Vladislav Golyanik","hidden":false},{"_id":"6a04bcd5b1a8cbabc9f08571","name":"Didier Stricker","hidden":false},{"_id":"6a04bcd5b1a8cbabc9f08572","name":"Alain Pagani","hidden":false}],"publishedAt":"2026-05-12T00:00:00.000Z","submittedOnDailyAt":"2026-05-13T00:00:00.000Z","title":"EgoForce: Forearm-Guided Camera-Space 3D Hand Pose from a Monocular Egocentric Camera","submittedOnDailyBy":{"_id":"641555d9385a75d7790fb343","avatarUrl":"/avatars/255aec71e94f380de5af809c766404b7.svg","isPro":true,"fullname":"Christen Milller","user":"chris10","type":"user","name":"chris10"},"summary":"Reconstructing the absolute 3D pose and shape of the hands from the user's viewpoint using a single head-mounted camera is crucial for practical egocentric interaction in AR/VR, telepresence, and hand-centric manipulation tasks, where sensing must remain compact and unobtrusive. While monocular RGB methods have made progress, they remain constrained by depth-scale ambiguity and struggle to generalize across the diverse optical configurations of head-mounted devices. As a result, models typically require extensive training on device-specific datasets, which are costly and laborious to acquire. This paper addresses these challenges by introducing EgoForce, a monocular 3D hand reconstruction framework that recovers robust, absolute 3D hand pose and its position from the user's (camera-space) viewpoint. EgoForce operates across fisheye, perspective, and distorted wide-FOV camera models using a single unified network. Our approach combines a differentiable forearm representation that stabilizes hand pose, a unified arm-hand transformer that predicts both hand and forearm geometry from a single egocentric view, mitigating depth-scale ambiguity, and a ray space closed-form solver that enables absolute 3D pose recovery across diverse head-mounted camera models. Experiments on three egocentric benchmarks show that EgoForce achieves state-of-the-art 3D accuracy, reducing camera-space MPJPE by up to 28% on the HOT3D dataset compared to prior methods and maintaining consistent performance across camera configurations. For more details, visit the project page at https://dfki-av.github.io/EgoForce.","upvotes":1,"discussionId":"6a04bcd5b1a8cbabc9f08573","projectPage":"https://dfki-av.github.io/EgoForce/","githubRepo":"https://github.com/dfki-av/EgoForce","githubRepoAddedBy":"user","ai_summary":"EgoForce is a monocular 3D hand reconstruction framework that uses a unified network to recover robust, absolute hand pose and position across different camera models through differentiable forearm representation, arm-hand transformers, and ray space solvers.","ai_keywords":["monocular 3D hand reconstruction","egocentric vision","camera-space viewpoint","differentiable forearm representation","arm-hand transformer","ray space closed-form solver","depth-scale ambiguity","fisheye camera model","perspective camera model","distorted wide-FOV camera model"],"githubStars":4,"organization":{"_id":"68e3c1b372607eeeeeaa9662","name":"dfki-av","fullname":"Augmented Vision","avatar":"https://cdn-avatars.huggingface.co/v1/production/uploads/64da1dd841f0e6c0e9638b67/k24siR3YWnto5sMIRAwu9.png"}},"canReadDatabase":false,"canManagePapers":false,"canSubmit":false,"hasHfLevelAccess":false,"upvoted":false,"upvoters":[{"_id":"641555d9385a75d7790fb343","avatarUrl":"/avatars/255aec71e94f380de5af809c766404b7.svg","isPro":true,"fullname":"Christen Milller","user":"chris10","type":"user"}],"acceptLanguages":["en"],"dailyPaperRank":0,"organization":{"_id":"68e3c1b372607eeeeeaa9662","name":"dfki-av","fullname":"Augmented Vision","avatar":"https://cdn-avatars.huggingface.co/v1/production/uploads/64da1dd841f0e6c0e9638b67/k24siR3YWnto5sMIRAwu9.png"},"markdownContentUrl":"https://huggingface.co/buckets/huggingchat/papers-content/resolve/2605/2605.12498.md"}">

Get this paper in your agent:

Don't have the latest CLI?

Add this paper to a collection to link it from this page.